settings not saved after update #477
Comments
|
is this perhaps an old value? the choices don't have that in the list |
I cah chose the same settings here? |
|
maybe you have to re-select it and save? weird how it would use the old format if not |
yeah that is what I did. |
|
so is it still broken after selecting a new one? |
nono just informing you that I had that selected before updating and after updating this error pop up and I could not save any other settings until I fixed this one first. |
sorry have to reopen. Since the last update we had the problem with the config not beeing the same anymore with the date which was a easy fix to just set the date again. But since then I got alot of Capture failed. The recording is still in the cache folder, and the json has no chapters. Only seem to be some streamer/folder but the all have the same permissions and before the recording worked fine. here are the logs I could find: |
|
which version of streamlink is packaged into the docker container now? |
|
Seems to be the latest version. |
|
i'll check my recordings if it happens, that log is very confusing |
I did not copy everything but those parts where errors happened. |
|
multiple streams recorded now and it seems fine, maybe it's the emoji and other stuff in the filename? |
This I changed after I noticed that they did not get recorded. Because the filenames did change in the update aswell. And I added the title just for fun I can remove it and see. |
|
are the captured mp4 files actually playable? |
Yeah only the filenames was the default. |
fount this: |
|
maybe something happened while it was converting or capturing? i'm at a loss here at the moment. unless someone else also has this problem it might be difficult to reproduce. i'll add more debugging info in mediainfo but that's about it |
|
could you check if there are more verbose errors with mediainfo on the dev branch? no need to capture anything new, just using one of the broken vods |
Can I change back after? I deleted the old ones because they where just taking away space. |
|
i would probably not recommend changing back in the same instance |
Then how would I see the old files if I don't use thos instance. Would have to setup a new one with dev and new config and cache ECT. So dunno how I should go about this. |
|
i don't know how those docker hosts work, but copying the data directory and pointing a new instance to it |
could I use the same storage directory because this is quite big. |
|
it might overwrite old versions with new stuff, unsure if i've added something new though. you could just copy the failing videos |
ok will try thanks for the fast responses. |
I hope I got everyything you wanted. |
|
i feel it must be the filename characters, unless the debug mode shows more |
|
hmm, try checking the capture logs for the failing streamers, both stdout and stderr. are you using an adblock proxy? |
No not using the AdBlock Proxy. Will check the logs later. |
the only things I could find in the logs are this: its almost like a virus it keeps going this was ok before so def not a recording thing my guess, this streamer is not online atm: |
|
wondering if it's some kind of network issue on your end, or something twitch is doing to block streaming... so it outputs to the ts files and fails, and the ts files are just partial streams, right? |
I doubt that its a network issue kinda. Everything else is working like intended. |
happening now to all channels: |
|
just so we're clear about this;
or
or what else is going on? still very confusing. |
It's neither. There are multiple streams being registered or somehow split I think. They are the same stream but are registered somehow. It looks like the first Json is being updated but the last is the one recording somehow. But the crazy part is that in my opinion it happens somewhere not directly when the stream starts. I checked the last days and it at first glance seems to be working fine, but when I check the next morning or something the got all fucked up |
|
cosmic rays? yeah this just makes no sense. would more lean on docker or the environment. i see nothing of this and there are no functions to save multiples of one vod |
|
and just as i said that i notice that two of my capturing streams stopped half an hour ago, this must be a streamlink issue |
Oh wow well maybe progress. |
it might have been a stream pause, not sure yet. still no broken streams anywhere |
Will report back after the update and a stream breaking again. |
I don't get it. Was working for one stream and now it's happening again. Internet was up the whole time have uptime Kuma running now. Space was sufficient the whole time and ram with 64gb cannot be the problem. Look at this specific stream yesterday an everything was working fine. And now it's worked again.
|
|
is the channel header/menu item for the streamer red/flashing/live? |
Nope
|
checked for another channel where the same problem occured yesterday, this is what I found in the log there seems to be some kind of pause and then it changes to no chapters found: and after that I looked even further back and found this? After that I thought this almost looks like it is rebooting somehow and found this: |
|
well shit, it's #444 i guess? hmm, wonder how i'd debug this |
Add one of the streamers mentioned and add more debugging logs? |
Dunno if this will help? https://stackoverflow.com/questions/38558989/node-js-heap-out-of-memory |
|
i've never gotten a memory leak on my server, so it would be difficult to create one |
|
i suppose it could be worth disabling watching storage in config |
you mean this one? |
|
yeah, it's the one thing that i know about that might affect memory. your logs say it's enabled so |
I enabled it just now as I posted the screen. So maybe a bit confusing. |
|
so it was disabled from the start? i do see entries in the logs saying it isn't... |
Like I said I enabled it before taking the screen. |
|
i noticed that the option did not disable all the watchers, the upcoming build should hopefully fix that |
For the last couple of days everything was working better than before. But now no stream is recording. All seem to have failed because something related to this: 2023-11-30T20:13:22.120Z | automator.fallbackCapture Spawned process 115 for fbcapture_bonjwa_41473701928 |
Seems like the date settings have changed and need to be updated before you can change the settings again:
The text was updated successfully, but these errors were encountered: